# Install the package from Bioconductor
if (!requireNamespace("BiocManager", quietly = TRUE)) {
install.packages("BiocManager")
}
BiocManager::install("Statial")# Loading required packages
library(Statial)
library(spicyR)
library(ClassifyR)
library(lisaClust)
library(dplyr)
library(SingleCellExperiment)
library(ggplot2)
library(ggsurvfit)
library(survival)
library(tibble)
library(treekoR)
theme_set(theme_classic())
nCores <- 1There are over 37 trillion cells in the human body, each taking up different forms and functions. The behaviour of these cells can be described by canonical characteristics, but their functions can also dynamically change based on their environmental context. Understanding the interplay between cells is key to understanding their mechanisms of action and how they contribute to human disease. Statial is a suite of functions to quantify the spatial relationships between cell types. This guide will provide a step-by-step overview of some key functions within Statial.
To illustrate the functionality of Statial we will use a multiplexed ion beam imaging by time-of-flight (MIBI-TOF) dataset profiling tissue from triple-negative breast cancer patients\(^1\) by Keren et al., 2018. This dataset simultaneously quantifies in situ expression of 36 proteins in 34 immune rich patients. Note: The data contains some “uninformative” probes and the original cohort included 41 patients.
These images are stored in a SingleCellExperiment object called kerenSCE. This object contains 57811 cells across 10 images and includes information on cell type and patient survival.
Note: The original dataset was reduced down from 41 images to 10 images for the purposes of this vignette, due to size restrictions.
# Load breast cancer
data("kerenSCE")
kerenSCE
#> class: SingleCellExperiment
#> dim: 48 57811
#> metadata(0):
#> assays(1): intensities
#> rownames(48): Na Si ... Ta Au
#> rowData names(0):
#> colnames(57811): 1 2 ... 171281 171282
#> colData names(8): x y ... Survival_days_capped Censored
#> reducedDimNames(0):
#> mainExpName: NULL
#> altExpNames(0):Kontextual is a method for performing inference on cell localisation which explicitly defines the contexts in which spatial relationships between cells can be identified and interpreted. These contexts may represent landmarks, spatial domains, or groups of functionally similar cells which are consistent across regions. By modelling spatial relationships between cells relative to these contexts, Kontextual produces robust spatial quantifications that are not confounded by biases such as the choice of region to image and the tissue structure present in the images.
In this example we demonstrate how cell type hierarchies can be used as a means to derive appropriate “contexts” for the evaluation of cell localisation. We then demonstrate the types of conclusions which Kontextual enables.
A cell type hierarchy may be used to define the “context” in which cell type relationships are evaluated within. A cell type hierarchy defines how cell types are functionally related to one another. The bottom of the hierarchy represents homogeneous populations of a cell type (child), the cell populations at the nodes of the hierarchy represent broader parent populations with shared generalised function. For example CD4 T cells may be considered a child population to the Immune parent population.
There are two ways to define the cell type hierarchy. First, they can be defined based on our biological understanding of the cell types. We can represent this by creating a named list containing the names of each parent and the associated vector of child cell types.
Note: The all vector must be created to include cell types which do not have a parent e.g. the undefined cell type in this data set.
# Examine all cell types in image
unique(kerenSCE$cellType)
#> [1] "Keratin_Tumour" "CD3_Cell" "B" "CD4_Cell"
#> [5] "Dc/Mono" "Unidentified" "Macrophages" "CD8_Cell"
#> [9] "other immune" "Endothelial" "Mono/Neu" "Mesenchymal"
#> [13] "Neutrophils" "NK" "Tumour" "DC"
#> [17] "Tregs"
# Named list of parents and their child cell types
biologicalHierarchy = list(
"tumour" = c("Keratin_Tumour", "Tumour"),
"tcells" = c("CD3_Cell", "CD4_Cell", "CD8_Cell", "Tregs"),
"myeloid" = c("Dc/Mono", "DC", "Mono/Neu", "Macrophages", "Neutrophils"),
"tissue" = c("Endothelial", "Mesenchymal")
)
# Adding more broader immune parent populationse
biologicalHierarchy$immune = c(biologicalHierarchy$bcells,
biologicalHierarchy$tcells,
biologicalHierarchy$myeloid,
"NK", "other immune", "B")
# Creating a vector for all cellTypes
all <- unique(kerenSCE$cellType)Alternatively, you can use the treeKor bioconductor package treekoR to define these hierarchies in a data driven way.
Note: These parent populations may not be accurate as we are using a small subset of the data.
# Calculate hierarchy using treekoR
kerenTree <- treekoR::getClusterTree(t(assay(kerenSCE, "intensities")),
kerenSCE$cellType,
hierarchy_method="hopach",
hopach_K = 1)
# Convert treekoR result to a name list of parents and children.
treekorParents = getParentPhylo(kerenTree)
treekorParents
#> $parent_1
#> [1] "Dc/Mono" "Macrophages" "NK"
#>
#> $parent_2
#> [1] "CD3_Cell" "B" "CD4_Cell" "CD8_Cell" "Tregs"
#>
#> $parent_3
#> [1] "Endothelial" "Mesenchymal" "DC"
#>
#> $parent_4
#> [1] "Unidentified" "other immune" "Mono/Neu" "Neutrophils" "Tumour"Here we examine an image highlighted in the Keren et al. 2018 manuscript where accounting for context information enables new conclusions. In image 6 of the Keren et al. dataset, we can see that p53+ tumour cells and immune cells are dispersed i.e. these two cell types are not mixing. However we can also see that p53+ tumour cells appear much more localised to immune cells relative to the tumour context (tumour cells and p53+ tumour cells).
# Lets define a new cell type vector
kerenSCE$cellTypeNew <- kerenSCE$cellType
# Select for all cells that express higher than baseline level of p53
p53Pos <- assay(kerenSCE)["p53", ] > -0.300460
# Find p53+ tumour cells
kerenSCE$cellTypeNew[kerenSCE$cellType %in% biologicalHierarchy$tumour] <- "Tumour"
kerenSCE$cellTypeNew[p53Pos & kerenSCE$cellType %in% biologicalHierarchy$tumour] <- "p53_Tumour"
# Group all immune cells under the name "Immune"
kerenSCE$cellTypeNew[kerenSCE$cellType %in% biologicalHierarchy$immune] <- "Immune"
# Plot image 6
kerenSCE |>
colData() |>
as.data.frame() |>
filter(imageID == "6") |>
filter(cellTypeNew %in% c("Immune", "Tumour", "p53_Tumour")) |>
arrange(cellTypeNew) |>
ggplot(aes(x = x, y = y, color = cellTypeNew)) +
geom_point(size = 1) +
scale_colour_manual(values = c("Immune" = "#505050", "p53_Tumour" = "#64BC46", "Tumour" = "#D6D6D6")) +
guides(colour = guide_legend(title = "Cell types", override.aes = list(size = 3)))